Single-Molecule Tracking Studies of Flow-Induced Microdomain

A glass coverslip (FisherFinest Premium; 25 × 25 mm2, 0.2 mm thick) and a rectangularly .... (25) The radius of cylindrical PEO domains was estimated...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCB

Single-Molecule Tracking Studies of Flow-Induced Microdomain Alignment in Cylinder-Forming Polystyrene−Poly(ethylene oxide) Diblock Copolymer Films Khanh-Hoa Tran-Ba, Daniel A. Higgins,* and Takashi Ito* Department of Chemistry, Kansas State University, 213 CBC Building, Manhattan, Kansas 66506-0401, United States S Supporting Information *

ABSTRACT: Flow-based approaches are promising routes to preparation of aligned block copolymer microdomains within confined spaces. An in-depth characterization of such nanoscale morphologies within macroscopically nonuniform materials under ambient conditions is, however, often challenging. In this study, single-molecule tracking (SMT) methods were employed to probe the flow-induced alignment of cylindrical microdomains (ca. 22 nm in diameter) in polystyrene−poly(ethylene oxide) diblock copolymer (PS-bPEO) films. Films of micrometer-scale thicknesses were prepared by overlaying a benzene solution droplet on a glass coverslip with a rectangular glass plate, followed by solvent evaporation under a nitrogen atmosphere. The microdomain alignment was quantitatively assessed from SMT data exhibiting the diffusional motions of individual sulforhodamine B fluorescent probes that preferentially partitioned into cylindrical PEO microdomains. Better overall microdomain orientation along the flow direction was observed near the substrate interface in films prepared at a higher flow rate, suggesting that the microdomain alignment was primarily induced by shear flow. The SMT data also revealed the presence of micrometer-scale grains consisting of highly ordered microdomains with coherent orientation. The results of this study provide insights into shear-based preparation of aligned cylindrical microdomains in block copolymer films from solutions within confined spaces.



INTRODUCTION Block copolymers (BCPs) are macromolecules that consist of two or more distinct homopolymer fragments connected via a covalent bond. Upon microphase separation, these polymers are known to self-assemble into ordered, periodic nanostructures of spherical, cylindrical, gyroid, and lamellar morphologies with characteristic dimensions of 5−50 nm.1 These morphologies and their dimensions are tunable by adjusting the volume fraction, chemical incompatibility, and lengths of the individual polymer fragments. BCPs have thus been extensively studied as a means to fabricate nanostructured monolithic materials for various applications including templates for nanomaterial synthesis,2,3 photolithographic masks,2−4 separator membranes for energy generation and storage,5,6 and sensing7 and separation media.3,4,7,8 For many of these applications, monolithic materials comprising closely packed, highly aligned anisotropic (i.e., cylindrical and lamellar) microdomains over a macroscopic size are desired. Unfortunately, BCP monoliths often consist of micrometer-scale “grains” formed from highly ordered cylindrical or lamellar microdomains.9,10 Such grain formation often limits the continuity of microdomains to micrometer-length scales due to different microdomain orientations in adjacent grains. So far, a number of methods have been explored to align anisotropic BCP microdomains along a desired direction. These often involve application of external fields during microphase separation.11 Among them, shear-based approaches have been widely explored for several decades. These approaches are © 2014 American Chemical Society

simple, applicable to large sample volumes, and suitable for various BCP systems in melt and solution.12 In addition, they provide metastable microdomain morphologies different from thermodynamic nanostructures in equilibrium.13 For example, Keller et al. successfully prepared a millimeter-scale, monolithic plug with well-ordered microdomains via the extrusion of molten polystyrene−polybutadiene−polystyrene (SBS) triblock copolymer through a circular die.14 Small-angle X-ray scattering (SAXS) measurements of the resulting plug revealed arrays of cylindrical polystyrene (PS) microdomains exhibiting nearcrystalline order. In a later study, Albalak and Thomas reported the fabrication of millimeter-thick SBS films comprising highly oriented PS microdomains over macroscopic length scales by using a roll-casting approach.15 The formation of well-ordered cylindrical and lamellar PS microdomains aligned along the rotation direction was revealed by SAXS and transmission electron microscopy (TEM) and explained by the hydrodynamic flow and gradual drying of a SBS solution between two rotating cylinders. As compared to a polymer melt, microdomain formation/alignment from a polymer solution takes place in a more complex manner due to the complicated phase behavior and significant volume change of the BCP solution due to solvent evaporation.15,16 These pioneering works demonstrated the fabrication of well-aligned BCP microReceived: July 28, 2014 Revised: August 31, 2014 Published: September 2, 2014 11406

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

domains by shear flow of a melt or solution within a confined space. Shear flow has also been used to align other nanostructures such as nanowires17,18 and to elongate/align cylindrical micelles19−23 within microfluidic channels. However, flow-induced alignment does not always offer macroscopically uniform BCP monoliths with perfectly aligned microdomains. Instead, a randomly distributed mixture of micrometer-scale grains consisting of microdomains aligned and misaligned in the flow direction is often obtained, as previously observed in polystyrene−polyisoprene diblock copolymer films using polarized optical microscopy (POM).9 Recently, single-molecule tracking (SMT) methods have been employed to probe the order and orientation of onedimensional (1D) microdomains in thin BCP films.24−26 By these methods, microdomain morphologies are visualized by tracking the diffusional motion of individual fluorescent probe molecules that preferentially partition into the cylindrical microdomains. Quantitative information on the microdomain alignment can be obtained through detailed analysis of 1D single-molecule trajectories using orthogonal regression methods.27 For instance, SMT was previously used by us to quantitatively assess the alignment of cylindrical microdomains induced by directional solvent vapor penetration (SVP) in polystyrene−poly(ethylene oxide) diblock copolymer (PS-bPEO) films.25 Upon SVP with 1,4-dioxane, 1D diffusion of fluorescent probe molecules along the solvent penetration direction was observed over millimeter-scale distances in the film plane and across micrometer-scale distances in the film thickness dimension, reflecting the presence of highly ordered, elongated PEO microdomains. In contrast, two-dimensional (2D) diffusion of probe molecules was observed in films treated with toluene or benzene vapor, suggesting inefficient microdomain alignment and elongation. In general, SMT offers several advantages over conventional methods (e.g., TEM,15 SAXS,14−16,28−31 and POM9,10,16) for characterization of nanostructured materials. These are based on its capabilities to acquire quantitative information on the temporal and nanoscale spatial heterogeneity of most samples under ambient conditions.32 Indeed, SMT provides a unique means to simultaneously and quantitatively assess the orientation and size of individual BCP microdomains, local microdomain order, and the diffusion of individual molecules incorporated into the microdomains.25,26 In this study, flow-induced alignment of cylindrical PEO microdomains in PS-b-PEO films was quantitatively assessed using SMT. Thin PS-b-PEO films (Figure 1a) were prepared under a nitrogen atmosphere by depositing its benzene solution on a glass coverslip and overlaying it with a second rectangular coverslip (Figure 1c), followed by subsequent solvent evaporation. The solution flowed through the gap between the two glass plates primarily based on the pressure applied by the upper coverslip. Microdomain alignment was investigated by recording the diffusional motions of individual fluorescent molecules (sulforhodamine B (SRB); Figure 1b) in the SMT experiments. The SRB probe molecules were observed to preferentially partition into cylindrical PEO microdomains.25 The morphologies and orientations of individual microdomains and the extent of their order were quantitatively assessed from single-molecule trajectories using orthogonal regression methods.27 Here, sandwiched PS-b-PEO films were prepared at different drying temperatures (T = 20, 40, and 60 °C) and PS-b-PEO concentrations (C = 20, 30, and 40% w/w) to explore the dependence of microdomain organization on these

Figure 1. Chemical structures of (a) PS-b-PEO and (b) sulforhodamine B (SRB). (c) Schematic illustration of the procedure used to prepare a sandwiched PS-b-PEO film: A droplet (12.5, 25, and 50 μL) of PS-b-PEO solution, placed on an underlying glass coverslip (top), was sandwiched by a rectangular glass plate to induce the directional flow of the solution via a pressure applied by the top piece (middle). The solution spread along the gap between the two substrates to form a uniform film (bottom). The film was dried at constant temperature (T = 20, 40, and 60 °C) under a N2 atmosphere for 4, 2, or 1 days prior to the SMT measurements. Each sample was measured at five different in-plane positions (red dots; bottom right) along the flow direction at z ≈ 1 μm. Here, x = 0 mm corresponds to the right edge of the droplet prior to placement of the glass plate. SMT videos were also acquired at different vertical positions (z = 1, 10, and 20−35 μm; bottom left).

parameters. Subsequently, SMT data in thin PS-b-PEO films prepared at different solution flow rates (v = 1, 3, and 5 mm/s) were measured at different distances from the film−substrate interface to explore the role of shear flow in inducing PEO microdomain alignment. Importantly, the SMT data revealed the presence of micrometer-scale grain morphologies in the PSb-PEO films and allowed the quantitative assessment of order and orientation for cylindrical PEO microdomains in individual grains.



EXPERIMENTAL PROCEDURES Chemicals and Materials. PS-b-PEO (PS, Mn = 42 000 g/ mol: PEO, Mn = 11 500 g/mol; PS volume fraction = 0.8; Mw/ Mn = 1.07) was purchased from Polymer Source and used as received. HPLC-grade benzene and ACS-grade SRB were purchased from Acros Organics and used without further purification. A glass coverslip (FisherFinest Premium; 25 × 25 mm2, 0.2 mm thick) and a rectangularly cut thinner glass slide 11407

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

1−4 days at 20−60 °C, whereas negligible molecular motion was observed in films after complete solvent removal at 230 °C under vacuum (9 consecutive frames in length were analyzed and discussed. Each trajectory was fitted to a straight line using orthogonal regression methods.25,27 1D, 2D, and immobile trajectories, reflecting 1D-diffusing, 2D-diffusing, and immobile molecules, were distinguished based on the variation in single-molecule positions across (σδ) and along (σR) the trajectories. Specifically, trajectories with σδ > 48 nm were classified as 2D, while those with σδ ≤ 48 nm and σR ≥ 77 nm were classified as 1D. Trajectories with σδ ≤ 48 nm and σR < 77 nm were identified as immobile. The threshold values for σδ and σR were estimated from Monte Carlo simulations of singlemolecule diffusion. These simulations employed a signal-tonoise (S/N) ratio and a diffusion coefficient similar to the experimental SMT data.25,27 The in-plane orientation of each 1D trajectory was discussed by determining the tilt angle, θ, of the fitted straight line relative to the solution flow direction (θ = 0°). The average in-plane orientation of 1D trajectories, θ̅, in a single grain was determined by fitting each θ-distribution to a Gaussian function. The in-plane order parameter of 1D trajectories, ⟨P⟩, in a single grain was determined from the standard deviation of the Gaussian fit, σ = ⟨Δθ⟩, using eq 1:20

(Goldseal cover glass; 22 × 7 mm2, 0.1 mm thick) were used as substrate and top piece, respectively. PS-b-PEO solutions prepared in benzene were adjusted to be 20, 30 and 40% w/w. A 200 nM SRB solution in methanol was added to a PS-b-PEO solution to yield nominally 5 nM dye (in the PEO microdomains) for film preparation. Sample Preparation. SRB-doped PS-b-PEO films were prepared in a nitrogen-filled glovebox (relative humidity 9 frames in length obtained from the videos shown in (a). (c) Histograms showing the number of 1D trajectories, N1D, at different tilt angles, θ, relative to the flow direction (θ = 0°) obtained from SMT data shown in (a, b). Average tilt angle, θ̅, and order parameter, ⟨P⟩ , values obtained by fitting the distributions to Gaussian curves (solid lines) are also given. (d) Plots showing the relationship between ⟨P⟩ and θ̅ measured for individual grains. The total number of grains, Ngrain, measured in each type of sample is shown in Table 1. Note that (d) (middle) shows the same data as Figure 3c (middle) and Figure 4c (bottom), reflecting the results of the same sample series.

Table 1. Effects of Drying Temperature on the Fractions of 1D/2D-Diffusing and Immobile Molecules, Average Tilt Angle, and Order Parameter from the 1D Trajectoriesa T (°C)

1D

2D

immobile

|θ | (deg)b

⟨P⟩b

NGrainc

20 40 60

0.18 ± 0.08 0.33 ± 0.13 0.30 ± 0.11

0.37 ± 0.15 0.22 ± 0.14 0.20 ± 0.14

0.45 ± 0.18 0.43 ± 0.16 0.50 ± 0.19

45 ± 23 22 ± 23 10 ± 12

0.94 ± 0.07 0.93 ± 0.11 0.96 ± 0.05

18 52 16

a Average and standard deviation from SMT data acquired at five different horizontal positions near the film−substrate interface (z ≈ 1 μm) in each of the three (T = 20 and 60 °C) or nine (T = 40 °C) films. The films were prepared from 30% w/w PS-b-PEO solution at a flow rate of v = 5 mm/s and dried for 4 days (T = 20 °C), 2 days (T = 40 °C), or 1 day (T = 60 °C) under a nitrogen atmosphere prior to the SMT measurements. bAverage and standard deviation of the parameters obtained from multiple grains. cThe total number of micrometer-scale grains obtained from the three or nine films (15 or 45 SMT videos, respectively). Note that the data of 40 °C (Table 1, middle) are the same as those of 30% w/w (Table 2, middle) and 5 mm/s (Table 3, bottom), reflecting the same sample series.

the presence of a larger flow-induced shear force.28,29 In addition, the deterioration of microdomain alignment was reported if the monoliths were dried slowly.15 Considering the requirement to use partially solvent swollen PS-b-PEO films for SMT measurements,25 we first studied the effects of drying temperature (T = 20, 40, and 60 °C) on microdomain alignment in PS-b-PEO films. Subsequently, we assessed microdomain alignment in films prepared at different PS-bPEO concentrations (C = 20, 30, and 40% w/w) to find a solution composition suitable for the film preparation method shown in Figure 1c. Then, we investigated microdomain alignment in PS-b-PEO films prepared at different flow rates (v = 1, 3, and 5 mm/s) by controlling the solution volume (see

Experimental Section). SMT data measured at different vertical positions (z = 1, 10, and 20−35 μm) in addition to the flowrate dependence permitted us to attribute microdomain alignment to shear flow during film preparation. Effects of Drying Temperatures on Microdomain Alignment. Figure 2a shows representative wide-field fluorescence images acquired in PS-b-PEO films that were prepared from a 30% w/w benzene solution at v = 5 mm/s and then dried at 20 °C for 4 days (top), 40 °C for 2 days (middle), or 60 °C for 1 day (bottom). Each wide-field image depicts the first frame of a 1000-frame SMT video. Portions of the videos (from 101st to 200th frames) are provided in the Supporting Information (Videos 1−3). These images clearly show 11409

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

Figure 3. (a) Representative single-molecule trajectories >9 frames in length measured in films prepared at different PS-b-PEO concentrations (C = 20, 30, and 40% w/w) near the film−substrate interface (z ≈ 1 μm). The arrows in the images indicate the flow direction of PS-b-PEO solution. (b) Histograms showing N1D at different θ obtained from SMT data shown in (a). (c) Plots showing the relationship between ⟨P⟩ and θ̅ measured for individual grains. The Ngrain measured in each type of sample is shown in Table 2. Note that the single-molecule trajectory data and histogram for C = 30% w/w (a, b) (middle) are different from those in Figures 2b,c (middle) and 4a,b (bottom)/5a,b (top) due to the use of a different sample prepared under the same conditions. However, (c) (middle) shows the same data as Figures 2d (middle) and 4c (bottom), reflecting the results of the same sample series.

Table 2. Effects of PS-b-PEO Concentration on the Fractions of 1D/2D-Diffusing and Immobile Molecules, Average Tilt Angle, and Order Parameter from the 1D Trajectoriesa C (% w/w)

1D

2D

immobile

|θ | (deg)b

⟨P⟩b

NGrainc

20 30 40

0.35 ± 0.10 0.33 ± 0.13 0.32 ± 0.06

0.26 ± 0.12 0.22 ± 0.14 0.32 ± 0.08

0.39 ± 0.13 0.43 ± 0.16 0.36 ± 0.09

31 ± 26 22 ± 23 34 ± 29

0.86 ± 0.20 0.93 ± 0.11 0.90 ± 0.11

17 52 15

Average and standard deviation from SMT data acquired at five different horizontal positions near the film−substrate interface (z ≈ 1 μm) in each of the three (C = 20 and 40% w/w) or nine (C = 30% w/w) films. The flow rate was 10, 5, and 1 mm/s for 20, 30, and 40% w/w PS-b-PEO solutions, respectively, due to the difference in solution viscosity. The films were dried at 40 °C under a nitrogen atmosphere for 2 days prior to the SMT measurements. bAverage and standard deviation of the parameters obtained from multiple grains. cThe total number of micrometer-scale grains obtained from the three or nine films (15 or 45 SMT videos, respectively). Note that the data of 30% w/w (Table 2, middle) are the same as those of 40 °C (Table 1, middle) and 5 mm/s (Table 3, bottom), reflecting the same sample series. a

individual fluorescence spots of diffraction-limited size that correspond to single SRB molecules. Figure 2b shows singlemolecule trajectories obtained from the SMT video data shown in Figure 2a. A large number of 1D trajectories oriented in the solution flow direction were found in the films dried at 40 and 60 °C (Figure 2b, middle and bottom), reflecting the diffusion of SRB molecules along 1D-aligned cylindrical PEO microdomains over micrometer-length scales. In contrast, curved trajectories and 1D trajectories tilted from the flow direction were observed in the film dried at 20 °C (Figure 2b, top), reflecting the presence of curved and misoriented microdomains, respectively. In addition to the 1D and curved trajectories, these SMT data depict spot-like trajectories originating from immobile

probe molecules that are adsorbed at the film−substrate interface or at the PS−PEO boundary regions or are incorporated into unelongated PEO microdomains. Furthermore, some of the 1D trajectories have larger widths, reflecting molecular diffusion within less-confined pathways.25 Here, the curved and wider 1D trajectories are defined as 2D trajectories. The trajectories obtained from multiple SMT videos were classified as 1D, 2D, and immobile ones (Table 1) using orthogonal regression methods as discussed previously (see Experimental Procedures).25,27 Table 1 shows that PS-b-PEO films dried at 20 °C exhibited a larger fraction of 2D trajectories than those dried at the higher temperatures (>99% confidence level, CL), as described above. The higher 2D trajectory ratio at T = 20 °C may reflect chain relaxation induced during slow 11410

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

Figure 4. (a) Representative single-molecule trajectories >9 frames in length measured in films prepared at different flow rates (v = 1, 3, and 5 mm/ s) near the film−substrate interface (z ≈ 1 μm). The arrows in the images indicate the flow direction of PS-b-PEO solution. (b) Histograms showing N1D at different θ obtained from SMT data shown in (a). (c) Plots showing the relationship between ⟨P⟩ and θ̅ measured for individual grains. The Ngrain measured in each type of sample is shown in Table 3. Note that the single-molecule trajectory data and histogram for v = 5 mm/s (a, b) (bottom) are different from those in Figures 2b,c (middle) and 3a,b (middle) due to the use of a different sample prepared under the same conditions. However, (c) (bottom) shows the same data as Figures 2d (middle) and 3c (middle), reflecting the results of the same sample series.

Table 3. Effects of Solution Flow Rate on the Fractions of 1D/2D-Diffusing and Immobile Molecules, Average Tilt Angle, and Order Parameter from the 1D Trajectoriesa v (mm/s)

1D

2D

immobile

|θ | (deg)b

⟨P⟩b

NGrainc

1 3 5

0.34 ± 0.14 0.36 ± 0.09 0.33 ± 0.13

0.17 ± 0.08 0.24 ± 0.06 0.22 ± 0.14

0.49 ± 0.15 0.40 ± 0.06 0.43 ± 0.16

47 ± 28 30 ± 25 22 ± 23

0.85 ± 0.19 0.92 ± 0.12 0.93 ± 0.11

19 18 52

Average and standard deviation from SMT data acquired at five different horizontal positions near the film−substrate interface (z ≈ 1 μm) in each of the three (v = 1 and 3 mm/s) or nine (v = 5 mm/s) films. Such films were prepared from PS-b-PEO solutions (C = 30% w/w) and then dried at T = 40 °C for 2 days prior to the SMT measurements. bAverage and standard deviation of the parameters obtained from multiple grains. cThe total number of micrometer-scale grains obtained from the three or nine films (15 or 45 SMT videos, respectively). Note that the data of 5 mm/s (Table 3, bottom) are the same as those of 40 °C (Table 1, middle) and 30% w/w (Table 2, middle), reflecting the same sample series. a

of two grains. Importantly, each of them comprises microdomains with a coherent orientation. Indeed, the singlemolecule trajectory data in Figure 2b shows two grains based on trajectories aligned in two different directions that may be connected to give curved trajectories. Unfortunately, the size of individual grains was challenging to estimate from the SMT videos because the boundaries between neighboring grains were made unclear by the relatively large spacing between adjacent trajectories due to the low probe concentration employed. Table 1 summarizes the average in-plane orientation of 1D trajectories, |θ |, observed for different individual grains in multiple samples. It is clear that films dried at higher T exhibited trajectory orientations closer to the flow direction, as represented by a smaller |θ | (>90% CL). Faster solvent evaporation at higher T facilitated “locking up” of the flow-

solvent evaporation, which leads to defect formation in gel-like BCP mesophases.28,29 Further quantitative assessment of the flow-induced microdomain alignment was explored only for 1D trajectories. Figure 2c shows histograms of the trajectory tilt angles, θ, of individual 1D trajectories obtained from the SMT data shown in Figure 2b. The histograms are clearly peaked near 0° for films dried at higher T, whereas the histogram for the one dried at 20 °C shows two distinct distributions that peak around −17° and 58°. The single peaks with narrow θ-distributions for the former films indicate the presence of single grains comprising highly ordered microdomains within the 16 × 16 μm 2 region. In addition, the peak near 0° indicates predominant microdomain alignment in the solution flow direction. In contrast, the observation of two distinct distributions for the film dried at 20 °C suggests the presence 11411

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

Figure 5. (a) Representative single-molecule trajectories >9 frames in length recorded at different distances from the film−substrate interface (z = 1, 10, and 20 μm) in a PS-b-PEO film. The film was prepared at v = 5 mm/s, C = 30% w/w and dried at T = 40 °C for 2 days. The arrows in the images indicate the flow direction of PS-b-PEO solution. (b) Histograms showing N1D at different θ obtained from SMT data shown in (a). (c) Plots showing the relationship between ⟨P⟩ and θ̅ measured for individual grains. The Ngrain measured in each type of sample is shown in Table 4. Note that the single-molecule trajectory data and histogram for z = 1 μm (a, b) (top) are the same as those in Figure 4a,b (bottom). However, (c) (top) shows different data than Figure 4c (bottom) due to a smaller number of data set that were recorded at different z.

Table 4. Fractions of 1D/2D-Diffusing and Immobile Molecules, Average Tilt Angle, and Order Parameter from the 1D Trajectories Recorded at Different Distances from the Film−Substrate Interfacea z (μm)

1D

2D

immobile

|θ | (deg)b

⟨P⟩b

NGrainc

1 10 20−35

0.32 ± 0.15 0.33 ± 0.15 0.32 ± 0.08

0.22 ± 0.09 0.19 ± 0.09 0.21 ± 0.05

0.46 ± 0.15 0.47 ± 0.16 0.48 ± 0.09

20 ± 21 34 ± 25 35 ± 27

0.92 ± 0.09 0.93 ± 0.08 0.91 ± 0.11

18 17 12

Average and standard deviation from SMT data acquired at different horizontal positions in three different films. These films were prepared from PS-b-PEO solutions (C = 30% w/w) at v = 5 mm/s and then dried at T = 40 °C for 2 days prior to the SMT measurements. A series of SMT data were collected at three different z for each film. In total, 15 (z = 1 μm), 13 (z = 10 μm), and 8 SMT videos (z = 20−35 μm) were recorded. b Average and standard deviation of the parameters obtained from multiple grains. cThe total number of micrometer-scale grains obtained from the three films. Note that the data at 1 μm (Table 4, top) are different from those at 5 mm/s (Table 3, bottom) due to the limited number of SMT videos recorded at different z. a

high ⟨P⟩ regardless of T. There is no clear correlation between ⟨P⟩ and θ̅. These results suggest the involvement of different mechanisms in microdomain orientation and ordering: The former was enhanced by shear flow and relaxed during solvent evaporation, whereas the latter reflected the agglomeration of cylindrical microdomains in the benzene solution and was affected by solvent evaporation from fairly concentrated solutions. These results revealed the importance of higher solvent evaporation rates on preserving the flow-induced alignment of 1D microdomains in BCP films. Effects of PS-b-PEO Concentration on Microdomain Alignment. Figure 3a depicts representative single-molecule trajectories in PS-b-PEO films prepared from benzene solutions of different polymer concentrations (C = 20, 30, and 40% w/w) at a fixed v (5 mm/s) and T (40 °C). The corresponding SMT

induced microdomain orientation prior to their relaxation, as reported by Albalak and Thomas.15 The width of the θ-distribution provides a means for quantitative assessment of the orientational order of the cylindrical microdomains in individual grains. Narrower distributions reflect higher microdomain order. Interestingly, the order parameter ⟨P⟩ obtained from the widths of the distributions in individual grains was, on average, similarly high regardless of T (Table 1). This observation indicates that the packing of the microdomains in each grain was retained during the relaxation of microdomain orientation upon solvent evaporation. Figure 2d depicts the relationship between ⟨P⟩ and θ̅ measured for individual grains. It is clear that the deviation of θ̅ from 0° was smaller at higher T, whereas many grains exhibited 11412

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

obtained at v = 5 mm/s. In addition, the trajectory order was slightly improved at higher v, as represented by the larger ⟨P⟩ (Table 3; >80% CL for data at v = 1 mm/s as compared to the others). These observations reveal enhanced microdomain alignment in the flow direction at higher v, as anticipated for shear-induced microdomain alignment.29,30 Flow-Induced Microdomain Alignment at Different Distances from the Film−Substrate Interface. One of the advantages of SMT for materials characterization is its capability to probe nanostructures across micrometer-scale depths.25 It is well-known that a larger flow-induced shear force is applied near the interface between two media due to the higher local shear velocity. SMT was thus used to investigate the z-dependence (depth profile) of the microdomain alignment in sandwiched PS-b-PEO films. Figure 5a depicts representative single-molecule trajectories in a PS-b-PEO film measured at different distances from the film−substrate interface (z = 1, 10, and 20 μm). The film was prepared from a C = 30% w/w benzene solution at v = 5 mm/s and T = 40 °C. The corresponding SMT videos are provided in the Supporting Information (Videos 10−12). Both Figure 5a and Table 4 show a similarly large number of 1D trajectories, regardless of z. However, the trajectory orientation was dependent on z: The orientations of 1D trajectories deviated from the flow direction at larger z, as supported by the corresponding histograms (Figure 5b). There are two grains in the trajectory data recorded at z = 10 μm (Figure 5a, middle), yielding peaks at θ̅ = 2° and = −62° in Figure 5b (middle). The larger deviation of trajectory orientation at larger z is also evident from the deviation of θ̅ from 0° in Figure 5c and the |θ | values in Table 4 (>90% CL for data at z = 1 μm as compared to the others). Interestingly, the trajectory order in individual grains was similarly high (⟨P⟩ ≈ 0.9) regardless of z (Table 4). These results suggest the involvement of different mechanisms behind microdomain ordering (the grain formation) and orientation, as discussed above. The z-dependent microdomain orientation can be qualitatively understood by considering a shear force active at the solution−substrate interface during film preparation. The flow of polymer solution during sample preparation was primarily driven by the pressure applied upon the placement of the rectangular glass plate over the polymer solution. Such a pressure-driven flow is expected to exhibit a parabolic flow profile due to the presence of stagnant boundary layers on the surfaces of the two substrates.34 In a parabolic flow profile, the local shear velocity, and thus the shear force, is greatest in the region closest to the solution−substrate interface, resulting in the microdomain orientation defined by the flow direction at smaller z. However, the flow-induced shear force was not strong enough to increase the population of cylindrical 1D microdomains, as indicated by the similar 1D fractions measured in samples at different z and v (Tables 3 and 4). The observation that ⟨P⟩ is not strongly dependent upon either v or z suggests that the microdomain order was predominantly determined by their agglomeration in solution rather than by shear forces active during sample preparation (vide supra). The z-dependence of microdomain alignment is in sharp contrast to our previous results on microdomain alignment of the same polymer via SVP.25 PEO microdomains were uniformly aligned in the penetration direction of 1,4-dioxane vapor across the entire film thickness. This is probably because the driving force of the microdomain alignment is the vapor concentration gradient across the film and is independent of the

videos are provided in the Supporting Information (Videos 4− 6). A number of 1D trajectories, reflecting the diffusion of SRB molecules within elongated PEO microdomains, were observed at the three C examined (Figure 3a). Indeed, the ratio of 1D trajectories was similar regardless of C (Table 2). In contrast, the microdomain orientation and order was strongly affected by C: Many trajectories in the film prepared at C = 30% w/w (middle) were oriented in the flow direction, whereas those in the other films exhibited significant tilting (C = 40% w/w) and curving (C = 20% w/w). The histograms of the θ-distribution of individual 1D trajectories (Figure 3b) support the observations in Figure 3a. Note that the histogram for C = 20% w/w exhibited a single, sharp distribution, apparently because it does not reflect the presence of the curved trajectories. The average orientation of 1D trajectories (θ̅) in each grain was close to zero for these histograms, reflecting microdomain orientation in the flow direction. However, the histogram for the film prepared from a solution of 40% w/w (bottom) shows a broader distribution of θ, giving a smaller ⟨P⟩. For multiple samples/grains, trajectory orientation in the flow direction was more reproducibly observed for a 30% w/w solution, as shown in Figure 3c, and by |θ | relatively closer to 0° as compared to the others (Table 2; >80% CL). In addition, 30% w/w solutions offered 1D trajectories with slightly higher order than 20% w/w solutions, as indicated by the larger ⟨P⟩ shown in Table 2 (>90% CL). These results indicate that the 30% w/w solution is the most suitable to prepare PS-b-PEO films with the desired microdomain orientation. The observation of better-aligned microdomains at the intermediate concentration can be explained by considering solution viscosity and flow rate. It is known that the shearinduced alignment is enhanced at higher flow rates of a more viscous solution.28,29 Here, the solution viscosity increases with increasing C. In contrast, the flow rate was lower at higher C due to higher viscosity because the films were prepared from a fixed solution volume (50 μL) using a rectangular upper glass slide of constant mass: v ≈ 1, 5, and 10 mm/s for 40, 30, and 20% w/w solutions, respectively. The 40% w/w solution did not provide a sufficiently high v, while the 20% w/w solution did not have a sufficiently high solution viscosity, resulting in less efficient shear-based microdomain alignment as compared to the 30% w/w solution. Effects of Solution Flow Rate on Microdomain Alignment. To discuss the influences of shear forces on microdomain alignment more clearly, it is essential to measure PS-b-PEO films prepared at a fixed solution viscosity (i.e., at a fixed C) with different flow rates (v). Figure 4a shows representative single-molecule trajectories in PS-b-PEO films prepared at a different v (1, 3, and 5 mm/s) from 30% w/w benzene solutions at a fixed T (40 °C). The flow rates were varied by adjusting the volumes of PS-b-PEO solutions. The corresponding SMT videos are provided in the Supporting Information (Videos 7−9). Figure 4a clearly shows that films prepared at higher v exhibited 1D trajectories better-oriented in the flow direction, as supported by the histograms (Figure 4b) showing a larger number of 1D trajectories with θ close to 0°. The percentage of 1D trajectories was independent of v (Table 3). Importantly, the improvement of trajectory orientation at v = 5 mm/s was reproducibly observed, as indicated by the narrower distribution in θ̅ about 0° (Figure 4c) and the |θ | values closer to 0° (Table 3; >70% CL) from multiple grains 11413

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B



distance from the film−substrate interface.25,35 In contrast, the shear-induced alignment is intrinsically defined by the shear flow of a solution containing phase-separated nanostructures, and thus its efficiency is higher in regions closer to the solution−solid interface, as reported for extruded BCP films.31 In addition, the solvent content of the solution used for the flow-induced approach is higher than that of the polymer film exposed to the solvent vapor. The higher solvent content may enhance grain formation during solvent evaporation due to the higher flexibility of the microdomains. Thus, though the experimental procedures are simpler, the flow-based method examined in this study may not be superior to the vapor penetration approach for fabricating well-aligned microdomain morphologies.

CONCLUSIONS In this study, SMT was used to systematically investigate the flow-induced alignment of cylindrical PEO microdomains in PS-b-PEO films prepared by sandwiching a polymer solution droplet between two glass plates. SMT could be used to quantitatively assess the alignment of individual microdomains as well as microdomain order in individual micrometer-scale grains from 1D single-molecule diffusion trajectories. The microdomain orientation in a PS-b-PEO film prepared from its solution could be controlled under a pressure-driven flow and was consistent with a shear-flow mechanism, as indicated by the flow rate dependence and depth profile of microdomain orientation. In contrast, the fraction of 1D microdomains and microdomain order in individual grains was less sensitive to the sample preparation conditions, probably because they reflected microdomain agglomeration morphologies in the relatively concentrated solutions. The presence of micrometer-scale grains was found as a possible obstacle to fabrication of macroscopically aligned microdomains from a BCP solution. Furthermore, optimization of drying temperature was needed to reduce the misorientation of microdomains during solvent evaporation. These results reveal that SMT provides a useful tool to quantitatively assess nanoscale and microscale morphologies in thin BCP films. ASSOCIATED CONTENT

S Supporting Information *

Wide-field fluorescence videos and videos for flow rate measurements. This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

(1) Bates, F. S.; Fredrickson, G. H. Block Copolymers - Designer Soft Materials. Phys. Today 1999, 52, 32−38. (2) Bang, J.; Jeong, U.; Ryu, D. Y.; Russell, T. P.; Hawker, C. J. Block Copolymer Nanolithography: Translation of Molecular Level Control to Nanoscale Patterns. Adv. Mater. 2009, 21, 4769−4792. (3) Olson, D. A.; Chen, L.; Hillmyer, M. A. Templating Nanoporous Polymers with Ordered Block Copolymers. Chem. Mater. 2008, 20, 869−890. (4) Hillmyer, M. A. Nanoporous Materials from Block Copolymer Precursors. Adv. Polym. Sci. 2005, 190, 137−181. (5) Hallinan, D. T., Jr.; Balsara, N. P. Polymer Electrolytes. Annu. Rev. Mater. Res. 2013, 43, 503−525. (6) Young, W.-S.; Kuan, W.-F.; Epps, T. H., III Block Copolymer Electrolytes for Rechargeable Lithium Batteries. J. Polym. Sci., Part B: Polym. Phys. 2014, 52, 1−16. (7) Ito, T. Block Copolymer-Derived Monolithic Polymer Films and Membranes Comprising Self-Organized Cylindrical Nanopores for Chemical Sensing and Separations. Chem.Asian J. 2014, DOI: 10.1002/asia.201402136. (8) Jackson, E. A.; Hillmyer, M. A. Nanoporous Membranes Derived from Block Copolymers: From Drug Delivery to Water Filtration. ACS Nano 2010, 4, 3548−3553. (9) Wang, H.; Newstein, M. C.; Chang, M. Y.; Balsara, N. P.; Garetz, B. A. Birefringence and Depolarized Light Scattering of an Ordered Block Copolymer Melt under Shear Flow. Macromolecules 2000, 33, 3719−3730. (10) Chastek, T. Q.; Lodge, T. P. Grain Shapes and Growth Kinetics during Self-Assembly of Block Copolymers. J. Polym. Sci., Part B: Polym. Phys. 2006, 44, 481−491. (11) Darling, S. B. Directing the Self-Assembly of Block Copolymers. Prog. Polym. Sci. 2007, 32, 1152−1204. (12) Hamley, I. W. Structure and Flow Behaviour of Block Copolymers. J. Phys.: Condens. Matter 2001, 13, 643−671. (13) Colby, R. H. Block Copolymer Dynamics. Curr. Opin. Colloid In. 1996, 1, 454−465. (14) Keller, A.; Pedemonte, E.; Willmouth, F. M. Macro-lattice from Segregated Amorphous Phases of a Three Block Copolymer. Nature 1970, 225, 538−540. (15) Albalak, R. J.; Thomas, E. L. Microphase Separation of Block Copolymer Solutions in a Flow Field. J. Polym. Sci., Part B: Polym. Phys. 1993, 31, 37−46. (16) Lodge, T. P.; Pudil, B.; Hanley, K. J. The Full Phase Behavior for Block Copolymers in Solvents of Varying Selectivity. Macromolecules 2002, 35, 4707−4717. (17) Huang, Y.; Duan, X.; Wei, Q.; Lieber, C. M. Directed Assembly of One-Dimensional Nanostructures into Functional Networks. Science 2001, 291, 630−633. (18) Su, B.; Wu, Y. C.; Jiang, L. The Art of Aligning OneDimensional (1D) Nanostructures. Chem. Soc. Rev. 2012, 41, 7832− 7856. (19) Trau, M.; Yao, N.; Kim, E.; Xia, Y. N.; Whitesides, G. M.; Aksay, I. A. Microscopic Patterning of Orientated Mesoscopic Silica through Guided Growth. Nature 1997, 390, 674−676. (20) Kirkeminde, A. W.; Torres, T.; Ito, T.; Higgins, D. A. Multiple Diffusion Pathways in Pluronic F127 Mesophases Revealed by Single Molecule Tracking and Fluorescence Correlation Spectroscopy. J. Phys. Chem. B 2011, 115, 12736−12743. (21) Ruhle, B.; Davies, M.; Lebold, T.; Brauchle, C.; Bein, T. Highly Oriented Mesoporous Silica Channels Synthesized in Microgrooves and Visualized with Single-Molecule Diffusion. ACS Nano 2012, 6, 1948−1960. (22) Park, S. C.; Ito, T.; Higgins, D. A. Single Molecule Tracking Studies of Flow-Aligned Mesoporous Silica Monoliths: Aging-Time Dependence of Pore Order. J. Phys. Chem. B 2013, 117, 4222−4230. (23) Xu, H.; Minter, C. J.; Nagasaka, S.; Ito, T.; Higgins, D. A. Elongation, Alignment, and Guided Electrophoretic Migration of dsDNA in Flow-Aligned Hexagonal F127 Gels. J. Phys. Chem. B 2014, 118, 4151−4159.





Article

AUTHOR INFORMATION

Corresponding Authors

*E-mail [email protected]; Tel 785-532-1451; Fax 785-532-6666 (T.I.). *E-mail [email protected]; Tel 785-532-6371; Fax 785-5326666 (D.A.H.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge Seok-Chan Park (Kansas State Univ.) for his help with the data analysis. They also acknowledge the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, of the U.S. Department of Energy (DE-FG02-12ER16095) for financial support of this work. 11414

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415

The Journal of Physical Chemistry B

Article

(24) Yorulmaz, M.; Kiraz, A.; Demirel, A. L. Motion of Single Terrylene Molecules in Confined Channels of Poly(butadiene)Poly(ethylene oxide) Diblock Copolymer. J. Phys. Chem. B 2009, 113, 9640−9643. (25) Tran-Ba, K.-H.; Finley, J. J.; Higgins, D. A.; Ito, T. SingleMolecule Tracking Studies of Millimeter-Scale Cylindrical Domain Alignment in Polystyrene-Poly(ethylene oxide) Diblock Copolymer Films Induced by Solvent Vapor Penetration. J. Phys. Chem. Lett. 2012, 3, 1968−1973. (26) Higgins, D. A.; Tran-Ba, K.-H.; Ito, T. Following Single Molecules to a Better Understanding of Self-Assembled OneDimensional Nanostructures. J. Phys. Chem. Lett. 2013, 4, 3095−3103. (27) Tran Ba, K. H.; Everett, T. A.; Ito, T.; Higgins, D. A. Trajectory Angle Determination in One Dimensional Single Molecule Tracking Data by Orthogonal Regression Analysis. Phys. Chem. Chem. Phys. 2011, 13, 1827−1835. (28) Pople, J. A.; Hamley, I. W.; Terrill, N. J.; Fairclough, J. P. A.; Ryan, A. J.; Yu, G.-E.; Booth, C. Shear-Induced Orientational Order in the Hexagonal Phase of Oxyethylene/Oxybutylene Diblock Copolymer Gels. Polymer 1998, 39, 4891−4896. (29) Pople, J. A.; Hamley, I. W.; Fairclough, J. P. A.; Ryan, A. J.; Booth, C. Orientational Ordering of a Poly(oxyethylene)-Poly(oxybutylene) Diblock Copolymer Gel under Steady Shear Flow. Macromolecules 1998, 31, 2952−2956. (30) Jain, A.; Hall, L. M.; Garcia, C. B. W.; Gruner, S. M.; Wiesner, U. Flow-Induced Alignment of Block Copolymer-Sol Nanoparticle Coassemblies toward Oriented Bulk Polymer-Silica Hybrids. Macromolecules 2005, 38, 10095−10100. (31) Shimizu, K.; Yasuda, T.; Saito, H. Perpendicular Orientation of Cylindrical Microdomains in Extruded Triblock Copolymer. Macromolecules 2010, 43, 2088−2091. (32) Moerner, W. E.; Fromm, D. P. Methods of Single-Molecule Fluorescence Spectroscopy and Microscopy. Rev. Sci. Instrum. 2003, 74, 3597−3619. (33) Nikoubashman, A.; Register, R. A.; Panagiotopoulos, A. Z. Simulations of Shear-Induced Morphological Transitions in Block Copolymers. Soft Matter 2013, 9, 9960−9971. (34) Whitesides, G. M.; Stroock, A. D. Flexible Methods for Microfluidics. Phys. Today 2001, 54, 42−48. (35) Osuji, C. O. Alignment of Self-Assembled Structures in Block Copolymer Films by Solvent Vapor Permeation. Macromolecules 2010, 43, 3132−3135.

11415

dx.doi.org/10.1021/jp507594t | J. Phys. Chem. B 2014, 118, 11406−11415